Optimization of laser irradiation process parameters for 40Cr steel based on back propagation neural network and genetic algorithm

Abstract

40Cr is a significant material in the manufacturing of gears. However, this process is plagued by uneven distribution of quenching layers and excessive residual stress, leading to a decline in the precision of the workpiece, deformation, and cracking. Relying solely on the trial-and-error method in experiments is inadequate for effectively revealing the evolution mechanism of the laser quenching process. This approach is detrimental to improving research and development efficiency. In this study, a multi-field coupled numerical model for the multi-track overlapping laser quenching of the 40Cr gear steel is established. This model quantitatively reveals the coupling evolution laws of the temperature field, stress field, and phase transformation field during the laser quenching. The focus is on calculating the size of the tempering zone under different overlapping rates and conducting correlation analysis between parameters. Based on the artificial neural networks and genetic algorithms, the width and depth of the tempering zone are selected as target values for parameter optimization. The aim is to find the nonlinear relationship between the laser power, spot diameter, scanning speed, laser overlapping rate, and the size of the tempering zone, which can accurately predict the size of the tempering zone. The material microstructure of 40Cr gear steel was characterized by the scanning electron microscope, microhardness tester and friction and wear tester. The results show that the training error and testing error of the output parameters in genetic algorithm optimized back propagation network can accurately predict the size of the tempering zone. It is found that the higher the lap rate, the more uniform the hardness of the phase transformation hardening layer surface, and the tempering softening zone becomes larger. The parts surface is easy to produce defects, which can provide a theoretical basis for optimizing the laser quenching process parameters.